Multiframe-based non-local means denoising for Raman spectra
Multiframe-based non-local means denoising for Raman spectra
复制标题
DOI：
10.1016/j.saa.2024.123931
复制
发表时间：
2024-01-31
影响因子：
4.4
通讯作者：
Zervas，Michalis N.
中科院分区：
文献类型：
--
作者：
Liu，Zhen;Ettabib，Mohamed A.;Zervas，Michalis N.
本文提出了一种拉曼光谱去噪的方法。该方法是基于这样的原则，即原始信号可以恢复平均像素的基础上结构相似性。相似性搜索和平均不限于相邻像素，而是扩展到跨不同帧的整个信号范围。该方法区别于传统的基于单帧相邻像素的滤波。通过在固定去噪参数下对模拟和实验拉曼数据进行去噪，验证了该方法的有效性和鲁棒性。几个去噪的结果和统计指标的模拟数据。从我们新开发的具有成本效益的波导增强拉曼光谱系统的实验拉曼光谱的恢复也被提出，并从传统的昂贵的拉曼显微镜相同的分析物的光谱相比。
A method for denoising Raman spectra is presented in this paper. The approach is based on the principle that the original signal can be restored by averaging pixels based on structure similarity. Similarity searching and averaging are not limited to the neighbouring pixels but extended throughout the entire signal range across different frames. This approach is distinguished from the conventional single-frame neighbour pixel-based filtering. The effectiveness and robustness of the proposed method are demonstrated through denoising simulated and experimental Raman data sets with fixed denoising parameters. Several denoised results and statistical indicators are presented for the simulated data. Recovery of the experimental Raman spectrum from our newly developed cost-effective waveguide-enhanced Raman spectroscopy system is also presented and compared to the spectrum from a conventional expensive Raman microscope for the same analyte.